SPSS Guidelines and Hypothesis Tests

SPSS Guidelines and Hypothesis Tests SPSS Guideline Obtaining the Covariance and Correlation

• Analyze → Correlate → Bivariate
• Drag your variables in the "Variables Box"

- For Covariance:

• Options → Cross-product deviations and covariance
• Click OK
• Read from the first variable in the table the covariance and correlation on the right in the

table 95% Confidence Intervals

• Analyze → Regression → Linear
• Determine the dependent variable and the independent variable(s)
• Click "Statistics" → "Confidence Intervals"
• Set Level
• Click "Continue" and "Ok"

Reading the ANOVA table Source Sum of Squares Degrees of Freedom Mean SquareF-value P-value (SPSS will state Sig.which is the P-value) RegressionSSR x x x x Residual SSE x x Total SST Reading the Coefficients table Variable Unstandardized B Standard Error Standardize d Coefficients t-value P-value (SPSS will state Sig.which is the P-value) Constant xx x x x B ₁ Regression Coefficient of B ₁ x x x x B ₂ Regression Coefficient of B ₂ x x x x 1 / 2

Prediction Intervals and Confidence Intervals

• For a certain value of a variable
• Type the value of the variable under the last observation
• Analyze → Regression → Linear → Determine variables
• Click "Save" and
• Tick "Unstandardized" under "Predicted Values"
• Tick "Mean" and "Individual" under "Prediction intervals"
• Set CI to 95%
• Click "Continue" and "Ok"  In data view you can see the interval

Definitions:

Mean Confidence Interval = LMCI and UMCI Individual Confidence Interval = LICI and UICI Directly Finding the Partial F-Test Statistic Value

• Analyze → Regression → Linear
• Determine the dependent variable and the independent variable(s) of the full model
• Click "Next"
• Drag the variables which are not in the reduced model in "Block 2 of 2"
• Select as method "Remove"
• Click "Statistics" → Tick "R squared change"

Model 1 = Full Model and Model 2 = Reduced Model Squared and Interaction Terms

For Square Term:

• Transform → Compute Variable → Name Variable

- Numeric Expression type: Variable * Variable

₁ ₁

For Interaction Term:

• Transform → Compute Variable → Name Variable

- Numeric Expression type: Variable * Variable

₁ ₂

• / 2